True Index: Domain Rankings

Private Top-Tier (S & A) standing across 3 core domains
Agentic & Enterprise90.7p
#5
4 Top-Tier private tests
Coding & SWE97.1p
#2
3 Top-Tier private tests
Math & Scientific Reasoning96.9p
#2
3 Top-Tier private tests

Head-to-Head: Clean Evals vs. Legacy Benchmarks

3 matched domains · positive boost means higher legacy scores
Capability Domain Clean Evals (Tiers S & A) Legacy Benchmarks (Tiers C & D) Public Lift (Boost)
Agentic & Enterprise90.7p (4)75.0p (2)-15.6p
Coding & SWE97.1p (3)93.3p (2)-3.8p
Math & Scientific Reasoning96.9p (3)97.1p (4)+0.2p
Clean Evals Avg (Tiers S & A)
95.3p
13 private, leak-resistant tests
Legacy Benchmarks Avg (Tiers C & D)
88.5p
8 public or aging tests
Matched Public Lift (Boost)
-6.6p
Near zero or negative indicates clean generalization
Machine-Readable Export

Audited Benchmark Performance of Claude Fable 5.2

Grouped by benchmark integrity so private capability signal stands apart from aging public tests.

Top-Tier Validated Tests (Tiers S & A) (13)

Private, leak-resistant tests that build real capability signal.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
FrontierCode-AA 2.0Private / FreshCoding & SWE#1/46100.0p
MKA (Multimodal Knowledge Agent)Private / FreshAgentic & Enterprise#1/23100.0p
MLCR-AAPrivate / FreshMath & Scientific Reasoning#1/46100.0p
FrontierMathPrivate / FreshMath & Scientific Reasoning#1/2799.5p
AA-Briefcase v1.1Private / FreshAgentic & Enterprise#3/5095.9p
AutomationBench-AAPrivate / FreshAgentic & Enterprise#5/5091.8p
Harvey LAB-AA v1.1Private / FreshAgentic & Enterprise#9/2175.0p
AA-Omniscience v1.1Private / FreshGeneral Knowledge & QA#1/47100.0p
AA-3DPrivate / FreshCoding & SWE#2/2298.8p
LiveBenchPrivate / FreshGeneral Knowledge & QA#4/2897.9p
Code Migration v1.1Private / FreshCoding & SWE#2/2492.6p
ARC-AGI-2Private / FreshGeneral Knowledge & QA#4/2291.9p
CritPt v2Private / FreshMath & Scientific Reasoning#1/4091.1p

Mid-Tier Evaluations (Tier B) (2)

Partially exposed or transitional evaluations with moderate leakage risk.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
HLE-DiamondSemi-ExposedGeneral Knowledge & QA#1/49100.0p
VIALS-AASemi-ExposedMath & Scientific Reasoning#1/23100.0p

Legacy & Vulnerable Benchmarks (Tiers C & D) (8)

Public or aging benchmarks prone to overfitting and training-data leaks.

BenchmarkBenchmark Integrity StatusDomainRankPercentile
BioMysteryBenchPublic / AgingMath & Scientific Reasoning#1/23100.0p
SWE Atlas - Codebase QnAPublic / AgingCoding & SWE#1/15100.0p
Terminal-Bench-Science 0.1Public / AgingMath & Scientific Reasoning#2/41100.0p
Autoform-HardPublic / AgingMath & Scientific Reasoning#2/894.9p
HealthBenchPublic / AgingMath & Scientific Reasoning#3/1693.5p
GDPval-AA v2.1Public / AgingAgentic & Enterprise#5/4386.8p
Terminal-Bench 4.0Public / AgingCoding & SWE#7/5086.7p
GDP.pdfPublic / AgingAgentic & Enterprise#19/5063.3p